Discursive Registers in Finno-Karelian Communicative Incantations

Abstract

The present paper investigates discursive justification as a moderative device in Kalevala-metric incantations. It explores different uses of justification by Finno-Karelian ritual specialists and argues that justification functions differently with positively and negatively evaluated non-human agents. In addition, a typology of discursive units of justification is provided. The results of the analysis reveal that justification functions as a register-based feature within the incantation genre. Types of justification directed to differently evaluated agents work as register-emblematic features. These findings open up new directions in the research of registers in Kalevala-metric incantation genre as constituted of multiple registers of communication.

Kalevala-metric incantations, a genre operating within a Finnic poetic system of Kalevala-meter, regularly invoke mythic non-human actors (e.g., Siikala 2002). The identity of these actors varies both within and between incantations recorded during 19^th and 20^th centuries from Finnish and Karelian ritual specialists called tietäjät (lit. ‘those who know, knowers’; sg. tietäjä). This paper sets out to investigate whether positively, negatively, and ambivalently evaluated actors are addressed with different discursive strategies. The analysis of possible variation in a performer’s strategies has potential to distinguish different registers within a genre, thus opening new directions for discussion about register-phenomena.

The starting hypothesis of the analysis is based on preliminary observations presented in an earlier article, where justificative means of Kalevala-metric incantations were explored (Karlsson 2021). The earlier research considered the use of moderative suffixes in incantations, and their possible role as indexes of moderation in directive discourse segments. It was found that suffixes were unlikely to play a significant role in indexing moderation towards positively evaluated non-human actors, and suffixal variation could be more feasibly accounted for as motivated by poetic meter. However, in the course of that investigation, discursive justification was observed to be connected in some cases with positively but not negatively evaluated agents, which led to the hypothesis of justification as a moderative device.

The examples of this type of justification are discursive acts that state the reason for directive statements (including requests and imperative forms), as in example (1), where justificative lines are underlined:

(1)

The present study explores the question of how discursive justification functions as a moderative instrument by examining the usage of justification in addressing different types of actors. It attempts to answer three questions:

1. 1) How are justificative segments in communicative incantations structured?

2. 2) Does discursive justification work as a moderative device in communicative incantations?

3. 3) Are expressions of justification organized into different types for differently evaluated actors?

The analysis below describes the organization of incantations into discourse units. I propose a preliminary typology of formal types of justifications and describe the conventions of use of these different types. The approach to the organization of communicative incantations draws on the concept of sequence organization from conversation analysis (see, e.g., Person 2016, 23–24). This converges with what Albert Lord (1960) broadly described as the ‘grammar’ of oral poetry (see also Foley 1995) and, when lines and groups of lines are in focus, with what has been described as ‘interlinear syntax’ in Kalevala-metric poetry (Frog 2016; see also Leino 1986, 129–161; Saarinen 2018).

Kalevala-metric poetry

Kalevala-metric poetry is a name for trochaic tetrameter, which was commonly used by the Finns, Karelians, Ingrians, Estonians, Votes and Ingrian-Finns as linguistically related ethnic groups (see Kuusi et al. 1977, 34–37) for various genres, both ritual and profane (Tarkka 2013, 53; 2016). Although the name anachronistically derives from the name of the Finnish national epic Kalevala (1835, extended edition in 1849; see Lönnrot 1999 and 2005), the oral-poetic system itself likely developed around 200–550 A.D. (Frog 2019; for discussion of the term Kalevala-meter, see Kallio 2011, 391; Kallio et al. 2017). In Viena Karelia (also called White Sea Karelia or Archangelsk Karelia), from which the material for this study originates, the poetic form was actively used until the early 20^th century, when it broke down in the wake of modernization (Tarkka 2013, 60).

A common feature of Kalevala-metric incantations is the effort to influence surrounding reality. Researchers have customarily categorized incantations into non-communicative or mechanical, and communicative incantations.¹ While non-communicative incantations were a widely used genre, communicative incantations were solely tools of the ritual specialists (Siikala 2002, Ch. III). Unlike the mechanical incantations, the incantations used by the tietäjät were attempts to establish communication with the otherworld through the use of directives and address terms for non-human actors (Siikala 2002, 84–86). Despite the categorization, it should be noted that so-called “non-communicative” incantations were probably also used by tietäjät without being mechanical,² and non-ritual specialist could use incantations that at least ostensibly addressed non-human agents (such as the forest in hunting incantations). The present paper focuses on attempts by tietäjät to communicate with other-worldly actors.

Research material

The data analyzed here is taken from the published edition of approximately 87,000 Kalevala-metric texts, Suomen Kansan Vanhat Runot, or SKVR (Ancient Poems of the Finnish People) (1908–1948, 1997), available in a digital edition that forms an electronically searchable corpus (skvr.fi).³ SKVR is organized by region and, within most regions, by genre. For this study, I have focused on the parishes of Viena Karelia, a region where collection efforts were extremely active, beginning from the second quarter of the nineteenth century (see map 1). Within this region, I developed a preliminary dataset of items categorized as incantations in SKVR from the parishes of Jyskyjärvi, Kieretti, Kiestinki, Kontokki, and Uhtua, constituting approximately 1,000 texts of varying length. These are all the parishes except for Vuokkiniemi. The decision to exclude Vuokkiniemi was mainly to keep the corpus manageable, since the parish boasts approximately 1,200 incantation texts published in SKVR, more than all the other parishes combined. In addition, the Kalevalaic poetry of Vuokkiniemi has been subject to comprehensive examination relatively recently, although that study has not addressed the type of variation on which I focus here (Tarkka 2005, 2013).

Figure 1. Viena Karelia on map

From the extant corpus of approximately 1,000 texts, I have delimited the material for this study by focusing on incantation texts in which directive utterances are present. I have thus omitted what I call “narrative incantations”: texts that do not include any directives and consist solely of third person narration.⁴ Example (2) illustrates the type of incantation that I exclude from the data to be analyzed. The text represents a variant of the Origin of Fire incantation, used by a tietäjä in healing performances. In this case, the text is probably incomplete: this example, collected in 1846, is from a time when collectors of the poetry were centrally interested in narrative poetry and especially lines and passages that they had not heard before. It was not uncommon for them only to record the portions of a poem that were directly of interest, and it is not unlikely that the collector of this text simply stopped writing down lines when the presentation advanced from narrating mythic events of the past to the ritual present. This text recounts how mythic heroes Väinämöinen and Joukahainen try to catch a fish that has swallowed an ember of fire that has fallen from heaven:

(2)

The incantation in (2) includes neither embedded speech nor the performer’s direct address to actors. The texts analyzed in this article, by contrast, are incantation texts that include both narrative and direct address.

The incantation in (3) is an example of this type, one connected to the healing of boils:

(3)

Limiting the data to incantations with directives reduces the initial corpus of approximately 1,000 texts to 515 incantations texts, which are used in this analysis. I focus on incantations that contain a performer’s speech to unseen agents.

Methodological considerations

In the discussion below, I examine patterned discursive forms that invoke images of stereotypical social personae and include voices indexing these personae and situations of interaction. The linguistic forms are then one aspect of discursive registers (Agha 2004, 2005, 2007; Blommaert 2007; Frog 2015; Kallio 2015; Visakko 2015). While discursive registers are often part of wider semiotic registers (Agha 2007, 80), the limitations of the available research material limit the scope of analysis to linguistic signs. Continuing the earlier investigation of moderative devices in incantations (Karlsson 2021), I am interested here in register-emblematic features of text-segments that consist of directives and justifications for them in incantations addressed to non-human actors.

In order to carry out the analysis, I first coded the data on the basis of the categories contained within discursive units. The categorization is based on thematic wholes, formed of single or multiple verses. The categorization is inspired by Anna-Leena Siikala’s work (1986; 2002: Ch. 4) on the content organization of different types of incantation but does not follow it directly. In her work, Siikala was interested in specific themes (like the banishment of illness, the call for non-human aides, and the naming of the adversary) that occur in incantations used by a tietäjä in addressing non-human actors. I am interested in how non-human actors are addressed, whether justification is used, and, if so, what types of justification are used for which types of addressee. Accordingly, I have applied the following classification (subscript lettering distinguishes subcategories):

1. A. Ambivalently attributing an actor⁶

2. C. Propitiating an actor

3. D. Directive utterance

   1. D_pos. Directive utterance to a positively-evaluated actor

   2. D_neg. Directive utterance to a negatively-evaluated actor

   3. D_amb. Directive utterance, evaluation of actor unclear

4. I. Inquiring utterance

5. J. Justificative utterance

6. N. Narrative discourse unit

   1. N_met. metanarrative discourse unit

7. V. Vocative utterance

A bar | is placed between utterances of different types, but utterances of the same type may be repeated in series, in which case the number of repetitions is indicated with a superscript number. Thus, the following passage in example (4) is described with the markup V│N³│V│D_amb│N, as it is organized as a vocative utterance (V), three narrative sequences (N³), a vocative utterance (V), and a directive addressed to an ambivalent agent (D_amb), followed by a final narrative section (N):

(4)

With narrative sections, which in the data most often describe surroundings or action without directly referring to an agent, the marking N_met is used to assign metanarrative units, where the speaker describes the incantation before the text itself or after it (example (5)):

(5)

All 515 incantations were analyzed in this way. The present study focuses on justifications that occur in 152 texts ranging from four to 212 lines in length. I note here that the units of more than one segment (e.g., I⁵) can be counted in the analysis as a single unit containing multiple tokens, or as the respective units of individual tokens in utterance units.

The smallest number of coded discourse units in the data is just one, distinguished in the shortest (four-line) text:

The largest is 50 (91 individual tokens) in the longest (212-line) text:

In the corpus of 515 analyzed texts, there is a total of 2445 discourse units (simplified numbering)⁸ and a total of approximately 3000 tokens (counting individual tokens, without grouping them). If not otherwise specified, simplified numbering is used in the discussion and examples that follow.

When identifying all justification tokens, a basic syntactic pattern of D + J was immediately apparent. The D + J sequence may either present J as a dependent clause or as an independent clause. When justifications regularly follow directives, the relative number of justifications to positively and negatively evaluated agents would remain unclear without clarifying the type of agent. This is the issue to which I now turn.

Status of actors and metapragmatic evaluations

The method of determining the status of an actor as positive, negative, or ambivalent, is based on metapragmatic evaluations, whether implicit or explicit. Implicit evaluations are marked by the way actors are addressed, and what they are asked to do. Explicit evaluations consist of verbal epithets, descriptions and assessments. The most common type of implicit evaluation in the data typifies the form of action: even when addressees are not named directly as positive or negative, what they are asked to do in the directive reveals their alignment with, or opposition to, the speaker. The example in (6) illustrates the implicit metapragmatic evaluation of an agent as positive, based on the action they are asked to perform and the desired result:

(6)

The attributes assigned to the actor in the excerpt (frosty, wearing an ice cap) are in themselves neutral. It is in the type of action requested that the evaluation is found: the agent is asked to bring cooling ice in order to help the tietäjä heal the injury of the patient. In some cases, explicit and implicit evaluation converge: in example (7), the interlocutor is addressed as Neitsyt Maria (Virgin Mary), and is regularly referred to explicitly as emonen (mother, lit. dam _diminutive) and rakas äiti armollinen (beloved merciful mother); these explicitly positive evaluations are followed by implicitly positive alignments. Here, the implicitly positively lines are the two final verses, where the performer describes the predicament of the woman or women giving birth (underlined):

(7)

The quantitative features of justification

The quantitative analysis of the data reveals that justification is not solely tied to directives addressed to positively evaluated agents. Justificative discourse units (J) occur in 152 incantation texts. In 58 incantations, D + J sequences are used exclusively to address positively evaluated actors so that the sequence is D_pos + J. In 22 incantations, the D + J sequence is used only to address a negatively evaluated non-human actor, or D_neg + J. In 15 incantations, the D + J sequence is directed towards both negatively and positively evaluated actors in the same text, i.e. D_pos + J and D_neg+ J and/or vice versa. In 18 incantations, the addressee is not treated as clearly positive or negative so that the sequence is D_amb + J. Finally, 39 texts include sequences, where D_amb is present in combination with either or both D_pos and D_neg.

On the level of tokens, 108 justificative units are directed to positive actors, 40 to ambivalent actors, and 38 to negative actors. The D_pos + J pattern constitutes the most substantial portion of D + J sequences, found in over twice the number of texts than D_neg + J, with almost three times as many tokens, while D_amb + J sequences appear in 18 texts with 40 total tokens, two more tokens than of the D_neg + J sequences.

Proportionately 40 % of the 267 texts with positive directives are connected with justification, while negative directives occur with justifications in 26 % of the 160 texts. With directives to ambivalent actors, the percentage is 31 % of 192 texts. On the token level, 23 % of the positive tokens are connected with justifications, while 15 % of the negative tokens are introduced with justification. In the ambivalent category D_amb, 14 % of the tokens are connected with justifications.

However, it will become apparent that justifications to positive, negative, and ambivalent agents are qualitatively different: justifications directed to positive agents frequently appear in the function of moderation. The justificative segments connected to negative agents, in turn, appeal to the benefit of the addressee; finally, justifications to ambivalent agents are founded on rationalizations that imply outcomes that would benefit both the performer and addressee. Moreover, justifications appear highly consistent inside categories, with few outliers. For analyzing the role of justifications as register-emblematic features, the qualitative difference between justifications offered to differently aligned non-human actors becomes a key area for investigation. I discuss this issue in more detail in the discussion below.

Types of justification

Type 1: Justifications to positive agents

The incantations with justificative segments address 16 different positive agents. Despite their diversity of identities, the acts they are asked to perform are formulated as having positive alignment with the speaker. Positive alignment means that the actors are asked to perform feats that directly aid the speaker or the (assumed) patient of the incantation: They are requested to come to the performer, to bring healing materials, or to put their hands on the patient, and so on (e.g., SKVR I₄, 318, 249, 1444). Regardless of the type of action, the ultimate result is beneficial to the performer and/or the patient of the ritual. Even when the agent is told to go somewhere else, as in the case of the honeybee, who brings ointments from distant mythic locations, they invariably return to the ritual site. On the level of justifications, this means that the justificative units focus on the needs of the speaker or patient, as in example (8), where a magical protection is sought:

(8)

Type 2: Justifications to negative agents

The negative actors exhibit a smaller variety of agent types: 10 in total. The acts they are asked to perform exhibit negative alignment with the ritual specialist and the human community. Seven out of ten of the negative actors are directed to move from the performer to some otherworldly location, often named as tarkka Tapiola (sharp Tapiola) or pimie Pohjola (dark Pohjola) (see Tarkka 2015, 23–28 about the attributes of the otherworld). This separates them clearly from the positive agents, who transfer themselves from the otherworld to the human sphere or move across the ritual borders. The justifications aimed at the negative actors prominently take two forms in the data. One of the segments consists of recounting how the mother (or, in some cases, the family) of the agent is waiting for and missing the malefactor at an otherworldly location. The other describes delicious offerings of food awaiting the culprit, such as “the best of roasts” in example (9):

(9)

It is noteworthy that the justification presented to negative actors does not appeal to a crisis on the part of the ritual specialist or patient, but focuses on the actor itself and its interests or desires.

Type 3: Justifications to ambivalent agents

The third type of justifications is connected with ambivalent agents, with five different actor types in the data. In this case, evaluations of actors formulate their liminal status: they are not treated as positive, but not directly negative either. A good example is the bear, an entity potentially harmful, but also serviceable as a means of livelihood. Interestingly, the common strategy of justification in this category is something that could be called rationalization. The justifications do not focus on crisis on behalf of the speaker or patient as a reason for the directive, nor do they directly order the recipient to move away from the ritual location, as with negative actors. Instead, the examples show that both parties would benefit from fulfilling the directive, as in example (10), where the performer makes an offer of peace for a bear:

(10)

Overview of the types of justification

The results of investigation of the justification types may not be very surprising: positive agents are addressed with different justifications than negative and ambivalent agents. However, the inner coherence of the types supports the argument that justification is a means of moderation when a performer uses directives with positive actors. Although justification also occurs in relation to negative directives, the relationship between the ritual specialist and the agent addressed is clearly different, showing negative actors as differently aligned relative to the performer. The justification segments connected with negative agents do not index moderation, but appeal instead to the agent’s concern for relatives or a desire for food. With ambivalent agents, the directives with justification fall into a category where neither moderation nor explicitly commanding discourse are present.

Non-human agents in the data are engaged according to regular schemas in which the evaluation is linked to the direction of the agent’s action relative to the performer. Direction of action describes movement or other ways of acting by the non-human actor, and its direction towards or away from the tietäjä. With positive actors, the direction is consistently towards the speaker: actors are requested and described as arriving to the aid of the performer. Correspondingly, negative actors are consistently told to go away from the ritual specialist (and the assumed patient) to locations described by the speaker. In addition, the specialist assumes different positions relative to the actors addressed. With positive agents, the relation is one of performer’s deference¹⁰ to addressee. When addressing negatively evaluated agents, the tietäjä takes a position of power and authority instead, issuing non-negotiable commands. Addresses to the ambivalent actors are from the same position, and the attributes given to the actors are neutral at best.

Conclusion

The above analysis explores the role of justification in incantations addressed to non-human actors. I have examined the question of how directive and justificative sequences differ for differently evaluated non-human actors. The results of the analysis demonstrate the existence of qualitative difference between the categories of D_pos + J, D_neg + J, and D_amb + J, which I summarize below.

The discursive sequence D_pos + J forms a category where a positively aligned non-human actor is addressed. Given the beneficial status of the actors addressed in such cases, the justification offered moderates the directive that accompanies it, a feature that is lacking in the other two categories. The actors in this category are consistently requested to perform actions, which aid the performer and the (assumed) patient. The contents of the justificative segments focus on describing the predicament of the speaker and their patients, and the performer places themselves in a position that requires assistance. I propose that this supports the hypothesis that justification can function as moderation when positive agents are addressed. While justifications also occur in D_neg + J and D_amb + J sequences, they do not function to express moderation in such cases.

The sequence D_neg + J is characterized by a discursive strategy where a negatively evaluated actor is ordered to depart from the performer. The directives are frequently qualitatively different from those expressed in the D_pos + J category, and include imperatives such as käsken (I command) (e.g., SKVR I₄, 578, 892, 894). Moreover, the justificative segments focus not on the good of the performer or the patient, but rather appeal to how the addressees’ kin or offerings of food await them in a remote mythic location. In this case, the performer assumes a more authoritative status that does not require the moderative justifications found in the case of positively evaluated agents.

Sequences of D_amb + J constitute the third type of justificative phenomenon. The status of non-human actors in this category is ambivalent, which results in the absence of both the moderation and the strong imperatives that we find in the D_pos + J and D_neg + J cases, respectively. In D_amb + J sequences, the justifications provide rationalizations for the directives, and formulate both the performer and addressee as beneficiaries.

These findings are satisfactorily explained in terms of register contrasts. The moderative use of justification with positively evaluated agents and the use of different types of justification with negative and ambivalent agents differentiates enregistered emblems (Agha 2007, 235) in incantation performance. The moderative use of justification indexes a positive alignment between the ritual specialist and non-human actors. The fact that this strategy is not used with negative and ambivalent agents shows that it is emblematic of a distinctive social relationship. The fact that distinct enregistered emblems are used in addressing different interlocutors in the incantation genre demonstrates the operation of register distinctions within a single discourse genre, and offers a new approach to the operation of register phenomena with Kalevala-metric poetry. This approach, which enables the investigation of different registers in Finno-Karelian communicative incantations can be extended and adjusted to other genres of Kalevala-metric poetry, and potentially to other oral traditions as well.